From bias to bots: discrimination and AI service preference
Li Min Lin, Youjae YiPurpose
While artificial intelligence (AI)-enabled services have become increasingly prevalent, little is known about how discriminatory service encounters influence consumers’ preference for AI-based service providers. This research aims to investigate whether – and why – perceived discrimination in human service encounters increases consumers’ preference for AI-driven service options.
Design/methodology/approach
Three studies were conducted to test the proposed framework. Study 1 used an autobiographical recall design, Study 2a experimentally manipulated discrimination, and Study 2b used a 2 × 2 between-subjects experimental design. Mediation and moderated mediation effects were examined using PROCESS.
Findings
Across the three studies, perceived discrimination was associated with greater preference for AI service providers through reduced social comfort with human providers and enhanced perceptions of AI ethicality. Cultural tightness consistently strengthened the social-comfort pathway, whereas its moderating influence on perceived AI ethicality and the direct effect was less consistent.
Practical implications
The findings offer actionable insights for designing inclusive service systems, deploying ethical AI solutions and enhancing customer experience strategies.
Originality/value
This research shows that human–AI service provider preference is shaped not only by the attributes of the available providers but also by consumers’ treatment in a preceding human-delivered service encounter. By linking discriminatory encounters to subsequent provider choice, the study extends research on service encounters, AI-enabled service and service inclusion and offers a socio-psychological explanation for why consumers may develop a stronger preference for AI-based service providers.